Starship Technologies is revolutionizing deliveries with autonomous robots. These robots are designed to deliver food, groceries, and packages across college campuses and neighborhoods in minutes. Starship has now completed millions of autonomous deliveries to date, traveled millions of miles, and is currently doing more than 140k road crossings each day. Our contribution to society includes reducing congestion, and pollution, providing zero-emissions deliveries, increasing the quality of life for residents, empowering seniors and disabled people, and enabling affordable delivery for local businesses.

We’re hitting records on an almost weekly basis. Now is an exciting time to join as we grow rapidly around the world!

We are looking for a Computer Vision Team Lead to join our Autonomous Driving division team in Estonia, Tallinn.

Our robots drive autonomously 99% of the time, with occasional help from remote human assistants. Computer vision is at the core of our robot's perceptual stack. Our algorithms operate in real time and span from low-level image processing to object detection, tracking, prediction, segmentation, and scene understanding using deep neural networks.

This is a hands-on role, where you’ll develop computer vision algorithms and neural networks for pushing Starship robots' autonomy level from 99% to 99.999%. In addition, you’ll lead the project to apply large neural networks to our robot autonomy, taking advantage of recent advances in the field as well as in inference hardware.

This position is located in Estonia, Tallinn. We are open to relocation for the right candidate.

Your main responsibilities are:

• Developing computer vision algorithms and neural networks to drive Starship robots' autonomy level from 99% to 99.999
• Leading the project to apply large neural networks to our robot autonomy, taking advantage of recent advances in the space as well as in inference hardware
• Designing neural networks for multiple platforms, including embedded boards, cloud GPUs, and cutting-edge hardware neural network accelerators
• Applying state-of-the-art multimodal LLMs to perception and corner cases comprehension in autonomous driving
• Designing processes and infrastructure for testing and active continuous improvement of neural networks
• Monitoring the accuracy and reliability of deployed modules on a large fleet of commercial-grade robots
• Making data-driven decisions for new algorithm development
• Leading a small team of engineers, guiding their work and supporting their development

What’s in it for you?

• You will collaborate with people who are passionate, motivated, and wonderfully capable. We are self-aware and seek feedback and improvement. We take pride in our ability to make the most of limited resources to solve problems no one has ever encountered before
  
• We are a fast-growing startup, with great opportunities to grow within the company. You will be part of a small team solving world-first problems at the forefront of autonomous driving at scale
  
• You’ll have endless opportunities to learn from our inspirational, talented team members across the globe, including some of the most experienced autonomous driving engineers in the world
  

What we hope you’ll bring to the table:

• Experience in developing and deploying complex neural network models
• Expertise in designing algorithms that meet strict performance requirements in real-time
• Ability to implement object detection and tracking solutions in practical scenarios
• Proficiency in C/C++/Rust, OpenCV, and Python
• Strong communication skills and ability to work well in a team
• Capability to work independently
• Self-driven and proactive attitude
• Team-leading experience or a strong interest in developing leadership skills

It would be a plus if you have familiarity with any of the following:

• Neural networks designed for embedded and real-time systems, such as YOLO, SSD, and MobileNet
• Practical applications of multimodal foundation models for solving complex problems
• Visual perception for autonomous driving
• Multi-sensory fusion
• Few-shot/semi-supervised/self-supervised learning
• GPU programming

Want to learn more about our robots and our people? Get in touch and let’s have a chat! Also, have a sneak peek to our blog and discover more about us!